The aim of this research was to investigate the possibility that there is a common susceptibility to false memories and the illusory truth effect that increases with age. The current study followed a quantitative quasi-experimental, within-subjects design, using three different age groups: young (16-39 years), middle-aged (40-60), and older adults (61+). Individuals participated in a number of pen and paper tests, including, eight false memory recognition tests using DRM (Deese/Roediger-McDermott) word lists, and an illusory truth effect task containing two sets of 50 statements.
Participants
Overall there were 161 participants (89 females and 72 males); their ages ranged from 16 to 92 years (mean 49.72, SD 19.57). The young age group consisted of 55
participants (34 females and 21 males), their average age was 27.47 years (SD = 6.39, range: 16-39 years). The middle-aged group had 53 participants (29 female and 24 male), their mean age was 50.36 years (SD = 5.94, range: 40-60 years). The older adult group contained 53 participants (26 female and 27 male), the groups mean age was 72.17 (SD = 7.92, range: 61-92 years). All participants were asked to use their hearing and reading aids if they needed to. Participants were volunteers, recruited via community groups and by word of mouth, from Marlborough, and the Palmerston North and wider Manawatu region.
Measures
The false recognition task. For this measure the DRM paradigm was employed. Eight lists, each consisting of 15 words, were used. Six lists were taken from Roediger and
McDermott’s (1995) original 24 lists, and two (smell and city) were from Stadler et al. (1999) that had been taken from unpublished work by McDermott (1995). Based on findings by Roediger and McDermott, and Stadler et al.’s follow-up research, and with the aim to
achieve a range of false memory scores, four lists found to reliably elicit the critical lure were chosen, alongside four lists moderately likely to elicit the critical lure. Both recall and
recognition scores from Stadler et al.’s research were used to examine the likelihood that each list would create a false memory because in their design recognition tests followed, and were likely influenced by, prior recall tests. Table 3.1, below, displays the eight DRM critical lures and list items used, along with the rankings given to each from Stadler et al.’s research.
Each DRM list had its own separate 15-item recognition test. Each test consisted of 11 list items, one critical lure, and three non-critical lures17
non-critical lures were used more than once in all of the eight tests. Lastly, all 15 numbers from DRM lists not used in the current research. Tests were constructed by randomly selecting, out of a hat, cards numbered 1-15. First the critical lure’s position in the test was established by picking a number. Then using random number generation on a Casio fx-82W the three non-critical lures were chosen, using only numbers 1-448 as each corresponded to one DRM item (including lures) of 28 non-used DRM lists stated in Stadler et al.’s (1999) appendix. Then three numbers were drawn from the hat to assign their position in the test. Note no
17 Non-critical lures are test items that were not presented at study, and were not semantically associated to
Table 3.1
DRM Critical Lures, List Items, and Their Rankings Based on Research by Stadler et al. (1999).
Rankingsa
Lure List Items Recall Recog.
Window Door, glass, pane, shade, ledge, sill, house, open, curtain, frame, view, breeze, sash, screen, shutter
1 1
Sleep Bed, rest, awake, tired, dream, wake, snooze, blanket, doze, slumber, snore, nap, peace, yawn, drowsy
2 7
Smell Nose, breathe, sniff, aroma, hear, see, nostril, whiff, scent, reek, stench, fragrance, perfume, salts, rose
3 2
Rough Smooth, bumpy, road, tough, sandpaper, jagged, ready, coarse, uneven, riders, rugged, sand, boards, ground, gravel
8 4
Mountain Hill, valley, climb, summit, top, molehill, peak, plain, glacier, goat, bike, climber, range, steep, ski
16 16
Slow Fast, lethargic, stop, listless, snail, cautious, delay, traffic, turtle, hesitant, speed, quick, sluggish, wait, molasses
17 17
River Water, stream, lake, Mississippi, boat, tide, swim, flow, run, barge, creek, brook, fish, bridge, winding
18 20
city Town, crowded, state, capital, streets, subway, country, New York, village, metropolis, big, Chicago, suburb, county, urban
13 22
Note. Recog. = Recognition.
a Rankings are based on results from 36 DRM lists ranked highest (#1) to lowest (#36) on their ability
were returned to the hat, this time each of the 15 numbers corresponded to one DRM list word (based on the order the items in each DRM list were written in Stadler et al.’s and Roediger and McDermott’s (1995) research). From the hat 11 numbers were haphazardly drawn and this determined which DRM words went in the test and their order (see Appendix B for all eight recognition tests).
In an attempt to control for order effects a balanced Latin square was used to organise the order in which participants heard the lists and completed the tests in the DRM task. This was used to ensure each list appeared before and after each other list an equal number of times, and each list had the opportunity to be an early, middle, and late item, in the presentation order. Table 3.2 below illustrates the eight list presentation conditions created using the balanced Latin square algorithm. Participants were randomly assigned to one of the conditions, with all conditions being used, approximately, an equal number of times across the three different age groups.
Table 3.2
The Eight DRM List Presentation Conditions, Ordered Using the Balanced Latin Square Algorithm
Condition DRM list presentation order
1 window, mountain, rough, sleep, river, city, smell, slow
2 mountain, rough, sleep, river, city, smell, slow, window
3 slow, window, mountain, rough, sleep, river, city, smell
4 rough, sleep, river, city, smell, slow, window, mountain
5 smell, slow, window, mountain, rough, sleep, river, city
6 sleep, river, city, smell, slow, window, mountain, rough
7 city, smell, slow, window, mountain, rough, sleep, river
8 river, city, smell, slow, window, mountain, rough, sleep
The illusory truth effect task. This task was adapted from Begg et al.’s (1992)
research in which 196 statements, each with a true and false version (created by modifying a detail such as a name or number) were used. Begg et al. labelled statements as true or false by using female and male sources (e.g., John Smith says toenails grow faster than
fingernails). To define true versus false sources, participants were told that statements said
by a female were true, and those said by a male were false, or vice versa. In the current study two different sets of 50 statements were created for the two different tasks (a
learning set and a test set). To make statements applicable to the New Zealand culture many of the American-oriented statements were eliminated, all imperial units were changed to their metric equivalents, and words not commonly used in New Zealand language were change to those that are (e.g., ‘the down of ten cashmere goats’ changed to ‘the wool of ten
cashmere goats’. In addition, when deciding to use either the true or false version of statements an attempt was made to evenly distribute longer and shorter statements between true and false categories. This was to avoid cueing participants to the truth of statements in the second illusory truth task (e.g., all false statements being long and true being short).
To create the learning set of statements (the first set of statements) all of Beggs et al.’s (1992) statements (except those eliminated as stated above) were placed in a box and 50 were haphazardly picked out. This determined which statements would be used and their presentation order. They were then divided into short and long statements, and again haphazardly divided so there was the same number of long and short statements in true and false categories. Female and male names were chosen using Van Overschelde, Rawson, and Dunlosky’s (2004) research on common female and male names. Names that could be considered appropriate for both genders (for example, Jamie, Alex, Pat) were excluded to avoid confusion and random number generation was used to pair names to statements.
The learning set of statements had two versions. One version (given to half the participants) had male names paired with true statements and female names with false statements. The second version (given to the other half of participants) was the same list of statements, in the same order, but female names were paired with true statements and male paired with false. This counterbalancing and confound check strategy was used in case there was a tendency for participants to view one gender as inherently more truthful than the other. In addition, next to each statement was a 6-point scale as the task required participants to indicate how interesting they thought each statement was (6 = very interesting, 1 = not interesting). See Appendix C for one version of the learning set of statements.
The test set of statements (the second set of statements) was developed to test participant’s memory for the truthfulness of the first set of statements, and again is based on Begg et al.’s (1992) study. The test set of statements contained 17 previously presented true statements (old true), 17 previously presented false statements (old false), and 16 statements that were not presented in the first set of statements (new); they were also drawn from Begg et al.’s list, randomly chosen in a haphazard fashion by drawing them from a box. Again, approximately the same number of long and short statements were used in each statement category. The 50 statements were then numbered 1-50 and random number generation was used to determine the order they appeared in the test. The only
specification for the order was that no more than three of the same type of statements (old true, old false, new) would appear one after the other. The test consisted of an instruction cover page, and six pages of statements. Next to each statement was a 6-point scale for participants to rate how true they thought each statement was (6 = certainly true, 1 = certainly false). See Appendix D for a copy of the test set of statements.
Apparatus
DRM word lists were read by a female and recorded onto a computer. Recording onto computer permitted precise timing of one word every 1.5 seconds. Each list was a separate mp3 audio file, and a bell sound indicated the end of each list. Lists were played at an appropriate volume depending on the size of each group, their physical spread, and hearing ability, through two Genius multimedia high-fidelity speakers connected to a Sansa mp3 player which was controlled during data collection by the researcher.
Procedure
Participants participated individually or in small groups of up to four. Because recruitment of participants was conducted in a variety of ways (stated earlier), a number of different sites were used to collect data. They included personal residences, community group meeting places, and participants’ workplaces. In all cases it was ensured that the room was quiet, private, and free from distractions.
Firstly, participants were asked to read an information sheet detailing the nature of the study (Appendix E) and sign an informed consent form (Appendix F). Following this, participants began the first task, which was the learning set of statements of the illusory truth task (Appendix C). Participants were verbally informed they would be given a set of 50 statements. They were asked to read each of these statements and rate each on the six point scale, by circling the number, to indicate how interesting they thought each statement was (six was ‘very interesting’ and one was ‘not interesting’). Depending on which version of the statements they were given they were told either that, statements said by a male are true, and those said by a female are false, or, statements said by a female are true, and those said by a male are false. In addition, participants were told the task had no time limit, and they could tear off the instructional cover page to keep in front of them to use as a reference while completing the task. They were also encouraged to ask questions if they needed further clarification. They were then given the learning set of statements, with the instructional cover page. This task took approximately 10-20 minutes for participants to complete.
Once participants completed reading and rating the learning set of statements the DRM task commenced. This task took approximately 10 minutes. For the DRM task, participants were told that they would hear eight lists of words, each made up of
approximately 15 words, played through speakers. They were told at the end of each list they would hear a bell sound, and after this sound the researcher would hand them a 15- word memory test (see Appendix B for the eight tests), on which they were to place a tick next to all the words they could remember from the previous list that was played. In addition, participants were told the test had no time limit, although it was not expected to take longer than two or three minutes. Participants were encouraged to ask any questions if they needed further clarification on the task. Once participants said they understood the task the first DRM list was played. After the bell sound each participant was given the
recognition test to complete. Once the test was completed, turned upside down, and moved to the side, the next DRM recording was played, followed by a recognition test. This
procedure continued until they had heard and completed recognition tests for all eight DRM lists.
After completion of the DRM tests, the third task involved the test set of the illusory truth effect statements (Appendix D). Participants were told they would be given another set of 50 statements, similar to those seen earlier. They were asked to read each of the
statements, and rate each on the six point scale, by circling the number, to indicate how true or false they thought it was (six was ‘certainly true’ one was ‘certainly false’). In addition, they were told that there were no ‘tricks’ involved – the researcher had not altered any details, such as names or dates, from the first set of statements to the second set to confuse them; the statement was either in the first set of statements or not. Participants were then told there was no time limit on the task, and they could tear off the instructional cover page to keep in front of them to use as a reference while completing the task. They were also encouraged to ask questions if they required further clarification on the process. When
participants were ready they were given the test set of statements with the instructional cover page. This task took approximately 10-20 minutes to complete.
The last task for participants was a simple questionnaire with two questions regarding age and identification as female or male (Appendix G). Participants were then asked if they knew what the purpose of the (DRM) word-list and recognition test was. This was to establish if participants had figured out that it was investigating if a critical lure was falsely remembered, which could alter their results. Only two people correctly guessed the purpose, and this was noted on their participant questionnaire with an asterisk. However, both felt it would have made no difference to their recognition results. The remainder of participants did not know what the task was investigating or thought that is was somehow related to the fact each list was made up of associated words.
Finally, because the research involved a low level of concealment, as participants were not specifically told that the study examines susceptibility to false memories and the illusory truth effect, participants were debriefed. Debriefing occurred immediately after data collection with each participant. The researcher first provided an explanation in lay terms of the types of memory ‘mistakes’ (the illusory truth effect and false memories in the DRM paradigm) that the project investigated and how they were investigated. Participants were also told this information was somewhat concealed because if they knew exactly what was being examined they may have been able use strategies to avoid making the mistakes. Participants were then encouraged to ask any questions they wished to have answered, or, voice any concerns, which were discussed. They were also reminded they could contact the researcher or supervisor at any time with further queries or concerns. Lastly, participants were asked not to tell other people about the information discussed during debriefing, or
the exact nature of the tasks, as these people may later be participants in the project and having them know this information could influence the results.
CHAPTER FOUR